tests: minor refactoring

[bison.git] / src / parse-gram.y

diff --git a/src/parse-gram.y b/src/parse-gram.y
index b897d6915fbd59b24a24d0bd2b52048d736bbf0b..0452c116943d6b799bd0f715759b6b838bdc9096 100644 (file)
--- a/src/parse-gram.y
+++ b/src/parse-gram.y
@@ -1,7 +1,6 @@
 %{/* Bison Grammar Parser                             -*- C -*-
 
 %{/* Bison Grammar Parser                             -*- C -*-
 

-   Copyright (C) 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010
-   Free Software Foundation, Inc.
+   Copyright (C) 2002-2013 Free Software Foundation, Inc.

 
    This file is part of Bison, the GNU Compiler Compiler.
 
 
    This file is part of Bison, the GNU Compiler Compiler.
 


@@ -21,6 +20,7 @@
 #include <config.h>
 #include "system.h"
 
 #include <config.h>
 #include "system.h"
 

+#include "c-ctype.h"

 #include "complain.h"
 #include "conflicts.h"
 #include "files.h"
 #include "complain.h"
 #include "conflicts.h"
 #include "files.h"


@@ -31,14 +31,16 @@
 #include "quotearg.h"
 #include "reader.h"
 #include "symlist.h"
 #include "quotearg.h"
 #include "reader.h"
 #include "symlist.h"

+#include "symtab.h"

 #include "scan-gram.h"
 #include "scan-code.h"
 #include "scan-gram.h"
 #include "scan-code.h"

+#include "xmemdup0.h"

 
 #define YYLLOC_DEFAULT(Current, Rhs, N)  (Current) = lloc_default (Rhs, N)
 static YYLTYPE lloc_default (YYLTYPE const *, int);
 
 #define YY_LOCATION_PRINT(File, Loc) \
 
 #define YYLLOC_DEFAULT(Current, Rhs, N)  (Current) = lloc_default (Rhs, N)
 static YYLTYPE lloc_default (YYLTYPE const *, int);
 
 #define YY_LOCATION_PRINT(File, Loc) \

-         location_print (File, Loc)
+          location_print (File, Loc)

 
 static void version_check (location const *loc, char const *version);
 
 
 static void version_check (location const *loc, char const *version);
 


@@ -46,7 +48,7 @@ static void version_check (location const *loc, char const *version);
    FIXME: depends on the undocumented availability of YYLLOC.  */
 #undef  yyerror
 #define yyerror(Msg) \
    FIXME: depends on the undocumented availability of YYLLOC.  */
 #undef  yyerror
 #define yyerror(Msg) \

-       gram_error (&yylloc, Msg)
+        gram_error (&yylloc, Msg)

 static void gram_error (location const *, char const *);
 
 static char const *char_name (char);
 static void gram_error (location const *, char const *);
 
 static char const *char_name (char);


@@ -57,24 +59,43 @@ static char const *char_name (char);
   static int current_prec = 0;
   static location current_lhs_location;
   static named_ref *current_lhs_named_ref;
   static int current_prec = 0;
   static location current_lhs_location;
   static named_ref *current_lhs_named_ref;

-  static symbol *current_lhs;
+  static symbol *current_lhs_symbol;

   static symbol_class current_class = unknown_sym;
   static uniqstr current_type = NULL;
 
   static symbol_class current_class = unknown_sym;
   static uniqstr current_type = NULL;
 

+  /** Set the new current left-hand side symbol, possibly common
+   * to several right-hand side parts of rule.
+   */
+  static
+  void
+  current_lhs(symbol *sym, location loc, named_ref *ref)
+  {
+    current_lhs_symbol = sym;
+    current_lhs_location = loc;
+    /* In order to simplify memory management, named references for lhs
+       are always assigned by deep copy into the current symbol_list
+       node.  This is because a single named-ref in the grammar may
+       result in several uses when the user factors lhs between several
+       rules using "|".  Therefore free the parser's original copy.  */
+    free (current_lhs_named_ref);
+    current_lhs_named_ref = ref;
+  }
+

   #define YYTYPE_INT16 int_fast16_t
   #define YYTYPE_INT8 int_fast8_t
   #define YYTYPE_UINT16 uint_fast16_t
   #define YYTYPE_UINT8 uint_fast8_t
 }
 
   #define YYTYPE_INT16 int_fast16_t
   #define YYTYPE_INT8 int_fast8_t
   #define YYTYPE_UINT16 uint_fast16_t
   #define YYTYPE_UINT8 uint_fast8_t
 }
 

-%debug
-%verbose
+%define api.prefix "gram_"
+%define api.pure
+%define locations
+%define parse.error verbose
+%define parse.lac full
+%define parse.trace

 %defines
 %defines

-%locations
-%pure-parser
-%define parse.error "verbose"
-%name-prefix="gram_"

 %expect 0
 %expect 0

+%verbose

 
 %initial-action
 {
 
 %initial-action
 {


@@ -118,7 +139,6 @@ static char const *char_name (char);
 %token PERCENT_DPREC         "%dprec"
 %token PERCENT_MERGE         "%merge"
 
 %token PERCENT_DPREC         "%dprec"
 %token PERCENT_MERGE         "%merge"
 

-

 /*----------------------.
 | Global Declarations.  |
 `----------------------*/
 /*----------------------.
 | Global Declarations.  |
 `----------------------*/


@@ -151,6 +171,7 @@ static char const *char_name (char);
 ;
 
 %token BRACED_CODE     "{...}"
 ;
 
 %token BRACED_CODE     "{...}"

+%token BRACED_PREDICATE "%?{...}"

 %token BRACKETED_ID    "[identifier]"
 %token CHAR            "char"
 %token EPILOGUE        "epilogue"
 %token BRACKETED_ID    "[identifier]"
 %token CHAR            "char"
 %token EPILOGUE        "epilogue"


@@ -166,30 +187,30 @@ static char const *char_name (char);
 %token TAG_NONE        "<>"
 
 %type <character> CHAR
 %token TAG_NONE        "<>"
 
 %type <character> CHAR

-%printer { fputs (char_name ($$), stderr); } CHAR
+%printer { fputs (char_name ($$), yyo); } CHAR

 
 /* braceless is not to be used for rule or symbol actions, as it
    calls code_props_plain_init.  */
 %type <chars> STRING "%{...%}" EPILOGUE braceless content.opt
 
 /* braceless is not to be used for rule or symbol actions, as it
    calls code_props_plain_init.  */
 %type <chars> STRING "%{...%}" EPILOGUE braceless content.opt

-%type <code> "{...}"
-%printer { fputs (quotearg_style (c_quoting_style, $$), stderr); }
-        STRING
-%printer { fprintf (stderr, "{\n%s\n}", $$); }
-        braceless content.opt "{...}" "%{...%}" EPILOGUE
-
-%type <uniqstr> BRACKETED_ID ID ID_COLON PERCENT_FLAG TAG variable
-%printer { fputs ($$, stderr); } <uniqstr>
-%printer { fprintf (stderr, "[%s]", $$); } BRACKETED_ID
-%printer { fprintf (stderr, "%s:", $$); } ID_COLON
-%printer { fprintf (stderr, "%%%s", $$); } PERCENT_FLAG
-%printer { fprintf (stderr, "<%s>", $$); } TAG
+%type <code> "{...}" "%?{...}"
+%printer { fputs (quotearg_style (c_quoting_style, $$), yyo); }
+         STRING
+%printer { fprintf (yyo, "{\n%s\n}", $$); }
+         braceless content.opt "{...}" "%{...%}" EPILOGUE
+
+%type <uniqstr> BRACKETED_ID ID ID_COLON PERCENT_FLAG TAG tag variable
+%printer { fputs ($$, yyo); } <uniqstr>
+%printer { fprintf (yyo, "[%s]", $$); } BRACKETED_ID
+%printer { fprintf (yyo, "%s:", $$); } ID_COLON
+%printer { fprintf (yyo, "%%%s", $$); } PERCENT_FLAG
+%printer { fprintf (yyo, "<%s>", $$); } TAG tag

 
 %type <integer> INT
 
 %type <integer> INT

-%printer { fprintf (stderr, "%d", $$); } <integer>
+%printer { fprintf (yyo, "%d", $$); } <integer>

 
 %type <symbol> id id_colon string_as_id symbol symbol.prec
 
 %type <symbol> id id_colon string_as_id symbol symbol.prec

-%printer { fprintf (stderr, "%s", $$->tag); } <symbol>
-%printer { fprintf (stderr, "%s:", $$->tag); } id_colon
+%printer { fprintf (yyo, "%s", $$->tag); } <symbol>
+%printer { fprintf (yyo, "%s:", $$->tag); } id_colon

 
 %type <assoc> precedence_declarator
 %type <list>  symbols.1 symbols.prec generic_symlist generic_symlist_item
 
 %type <assoc> precedence_declarator
 %type <list>  symbols.1 symbols.prec generic_symlist generic_symlist_item


@@ -252,9 +273,9 @@ input:
 ;
 
 
 ;
 
 

-       /*------------------------------------.
-       | Declarations: before the first %%.  |
-       `------------------------------------*/
+        /*------------------------------------.
+        | Declarations: before the first %%.  |
+        `------------------------------------*/

 
 prologue_declarations:
   /* Nothing */
 
 prologue_declarations:
   /* Nothing */


@@ -294,9 +315,8 @@ prologue_declaration:
                                     MUSCLE_PERCENT_DEFINE_GRAMMAR_FILE);
     }
 | "%expect" INT                    { expected_sr_conflicts = $2; }
                                     MUSCLE_PERCENT_DEFINE_GRAMMAR_FILE);
     }
 | "%expect" INT                    { expected_sr_conflicts = $2; }

-| "%expect-rr" INT                { expected_rr_conflicts = $2; }
+| "%expect-rr" INT                 { expected_rr_conflicts = $2; }

 | "%file-prefix" STRING            { spec_file_prefix = $2; }
 | "%file-prefix" STRING            { spec_file_prefix = $2; }

-| "%file-prefix" "=" STRING        { spec_file_prefix = $3; } /* deprecated */

 | "%glr-parser"
     {
       nondeterministic_parser = true;
 | "%glr-parser"
     {
       nondeterministic_parser = true;


@@ -311,13 +331,11 @@ prologue_declaration:
       muscle_code_grow ("initial_action", action.code, @2);
       code_scanner_last_string_free ();
     }
       muscle_code_grow ("initial_action", action.code, @2);
       code_scanner_last_string_free ();
     }

-| "%language" STRING           { language_argmatch ($2, grammar_prio, @1); }
+| "%language" STRING            { language_argmatch ($2, grammar_prio, @1); }

 | "%name-prefix" STRING         { spec_name_prefix = $2; }
 | "%name-prefix" STRING         { spec_name_prefix = $2; }

-| "%name-prefix" "=" STRING     { spec_name_prefix = $3; } /* deprecated */

 | "%no-lines"                   { no_lines_flag = true; }
 | "%no-lines"                   { no_lines_flag = true; }

-| "%nondeterministic-parser"   { nondeterministic_parser = true; }
+| "%nondeterministic-parser"    { nondeterministic_parser = true; }

 | "%output" STRING              { spec_outfile = $2; }
 | "%output" STRING              { spec_outfile = $2; }

-| "%output" "=" STRING          { spec_outfile = $3; }  /* deprecated */

 | "%param" { current_param = $1; } params { current_param = param_none; }
 | "%require" STRING             { version_check (&@2, $2); }
 | "%skeleton" STRING
 | "%param" { current_param = $1; } params { current_param = param_none; }
 | "%require" STRING             { version_check (&@2, $2); }
 | "%skeleton" STRING


@@ -335,7 +353,7 @@ prologue_declaration:
             xmalloc (dir_length + 1 + strlen (skeleton_user) + 1);
           if (dir_length > 0)
             {
             xmalloc (dir_length + 1 + strlen (skeleton_user) + 1);
           if (dir_length > 0)
             {

-              strncpy (skeleton_build, current_file, dir_length);
+              memcpy (skeleton_build, current_file, dir_length);

               skeleton_build[dir_length++] = '/';
             }
           strcpy (skeleton_build + dir_length, skeleton_user);
               skeleton_build[dir_length++] = '/';
             }
           strcpy (skeleton_build + dir_length, skeleton_user);


@@ -367,19 +385,17 @@ grammar_declaration:
     {
       grammar_start_symbol_set ($2, @2);
     }
     {
       grammar_start_symbol_set ($2, @2);
     }

-| "%destructor" "{...}" generic_symlist
+| code_props_type "{...}" generic_symlist

     {
     {

-      symbol_list *list;
-      for (list = $3; list; list = list->next)
-       symbol_list_destructor_set (list, $2, @2);
-      symbol_list_free ($3);
-    }
-| "%printer" "{...}" generic_symlist
-    {
-      symbol_list *list;
-      for (list = $3; list; list = list->next)
-       symbol_list_printer_set (list, $2, @2);
-      symbol_list_free ($3);
+      code_props code;
+      code_props_symbol_action_init (&code, $2, @2);
+      code_props_translate_code (&code);
+      {
+        symbol_list *list;
+        for (list = $3; list; list = list->next)
+          symbol_list_code_props_set (list, $1, &code);
+        symbol_list_free ($3);
+      }

     }
 | "%default-prec"
     {
     }
 | "%default-prec"
     {


@@ -403,6 +419,13 @@ grammar_declaration:
     }
 ;
 
     }
 ;
 

+%type <code_type> code_props_type;
+%union {code_props_type code_type;};
+%printer { fprintf (yyo, "%s", code_props_type_string ($$)); } <code_type>;
+code_props_type:
+  "%destructor"  { $$ = destructor; }
+| "%printer"     { $$ = printer; }
+;

 
 /*---------.
 | %union.  |
 
 /*---------.
 | %union.  |


@@ -443,7 +466,7 @@ symbol_declaration:
       symbol_list *list;
       tag_seen = true;
       for (list = $3; list; list = list->next)
       symbol_list *list;
       tag_seen = true;
       for (list = $3; list; list = list->next)

-       symbol_type_set (list->content.sym, $2, @2);
+        symbol_type_set (list->content.sym, $2, @2);

       symbol_list_free ($3);
     }
 ;
       symbol_list_free ($3);
     }
 ;


@@ -454,10 +477,10 @@ precedence_declaration:
       symbol_list *list;
       ++current_prec;
       for (list = $3; list; list = list->next)
       symbol_list *list;
       ++current_prec;
       for (list = $3; list; list = list->next)

-       {
-         symbol_type_set (list->content.sym, current_type, @2);
-         symbol_precedence_set (list->content.sym, current_prec, $1, @1);
-       }
+        {
+          symbol_type_set (list->content.sym, current_type, @2);
+          symbol_precedence_set (list->content.sym, current_prec, $1, @1);
+        }

       symbol_list_free ($3);
       current_type = NULL;
     }
       symbol_list_free ($3);
       current_type = NULL;
     }


@@ -484,9 +507,9 @@ symbols.prec:
 ;
 
 symbol.prec:
 ;
 
 symbol.prec:

-    symbol { $$ = $1; }
-  | symbol INT { $$ = $1; symbol_user_token_number_set ($1, $2, @2); }
-  ;
+  symbol     { $$ = $1; }
+| symbol INT { $$ = $1; symbol_user_token_number_set ($1, $2, @2); }
+;

 
 /* One or more symbols to be %typed. */
 symbols.1:
 
 /* One or more symbols to be %typed. */
 symbols.1:


@@ -503,9 +526,13 @@ generic_symlist:
 
 generic_symlist_item:
   symbol    { $$ = symbol_list_sym_new ($1, @1); }
 
 generic_symlist_item:
   symbol    { $$ = symbol_list_sym_new ($1, @1); }

-| TAG       { $$ = symbol_list_type_new ($1, @1); }
-| "<*>"     { $$ = symbol_list_default_tagged_new (@1); }
-| "<>"      { $$ = symbol_list_default_tagless_new (@1); }
+| tag       { $$ = symbol_list_type_new ($1, @1); }
+;
+
+tag:
+  TAG
+| "<*>" { $$ = uniqstr_new ("*"); }
+| "<>"  { $$ = uniqstr_new (""); }

 ;
 
 /* One token definition.  */
 ;
 
 /* One token definition.  */


@@ -548,9 +575,9 @@ symbol_defs.1:
 ;
 
 
 ;
 
 

-       /*------------------------------------------.
-       | The grammar section: between the two %%.  |
-       `------------------------------------------*/
+        /*------------------------------------------.
+        | The grammar section: between the two %%.  |
+        `------------------------------------------*/

 
 grammar:
   rules_or_grammar_declaration
 
 grammar:
   rules_or_grammar_declaration


@@ -569,8 +596,11 @@ rules_or_grammar_declaration:
 ;
 
 rules:
 ;
 
 rules:

-  id_colon named_ref.opt { current_lhs = $1; current_lhs_location = @1;
-    current_lhs_named_ref = $2; } rhses.1
+  id_colon named_ref.opt { current_lhs ($1, @1, $2); } rhses.1
+  {
+    /* Free the current lhs. */
+    current_lhs (0, @1, 0);
+  }

 ;
 
 rhses.1:
 ;
 
 rhses.1:


@@ -581,12 +611,14 @@ rhses.1:
 
 rhs:
   /* Nothing.  */
 
 rhs:
   /* Nothing.  */

-    { grammar_current_rule_begin (current_lhs, current_lhs_location,
-                                 current_lhs_named_ref); }
+    { grammar_current_rule_begin (current_lhs_symbol, current_lhs_location,
+                                  current_lhs_named_ref); }

 | rhs symbol named_ref.opt
     { grammar_current_rule_symbol_append ($2, @2, $3); }
 | rhs "{...}" named_ref.opt
 | rhs symbol named_ref.opt
     { grammar_current_rule_symbol_append ($2, @2, $3); }
 | rhs "{...}" named_ref.opt

-    { grammar_current_rule_action_append ($2, @2, $3); }
+    { grammar_current_rule_action_append ($2, @2, $3, false); }
+| rhs "%?{...}"
+    { grammar_current_rule_action_append ($2, @2, NULL, true); }

 | rhs "%prec" symbol
     { grammar_current_rule_prec_set ($3, @3); }
 | rhs "%dprec" INT
 | rhs "%prec" symbol
     { grammar_current_rule_prec_set ($3, @3); }
 | rhs "%dprec" INT


@@ -601,7 +633,6 @@ named_ref.opt:
   BRACKETED_ID   { $$ = named_ref_new($1, @1); }
 ;
 
   BRACKETED_ID   { $$ = named_ref_new($1, @1); }
 ;
 

-

 /*---------------------------.
 | variable and content.opt.  |
 `---------------------------*/
 /*---------------------------.
 | variable and content.opt.  |
 `---------------------------*/


@@ -690,7 +721,6 @@ epilogue.opt:
 
 %%
 
 
 %%
 

-

 /* Return the location of the left-hand side of a rule whose
    right-hand side is RHS[1] ... RHS[N].  Ignore empty nonterminals in
    the right-hand side, and return an empty location equal to the end
 /* Return the location of the left-hand side of a rule whose
    right-hand side is RHS[1] ... RHS[N].  Ignore empty nonterminals in
    the right-hand side, and return an empty location equal to the end


@@ -707,14 +737,14 @@ lloc_default (YYLTYPE const *rhs, int n)
   loc.start = rhs[n].end;
   loc.end = rhs[n].end;
 
   loc.start = rhs[n].end;
   loc.end = rhs[n].end;
 

-  /* Ignore empty nonterminals the start of the the right-hand side.
+  /* Ignore empty nonterminals the start of the right-hand side.

      Do not bother to ignore them at the end of the right-hand side,
      since empty nonterminals have the same end as their predecessors.  */
   for (i = 1; i <= n; i++)
     if (! equal_boundaries (rhs[i].start, rhs[i].end))
       {
      Do not bother to ignore them at the end of the right-hand side,
      since empty nonterminals have the same end as their predecessors.  */
   for (i = 1; i <= n; i++)
     if (! equal_boundaries (rhs[i].start, rhs[i].end))
       {

-       loc.start = rhs[i].start;
-       break;
+        loc.start = rhs[i].start;
+        break;

       }
 
   return loc;
       }
 
   return loc;


@@ -742,28 +772,20 @@ add_param (param_type type, char *decl, location loc)
 
     /* Strip the surrounding '{' and '}', and any blanks just inside
        the braces.  */
 
     /* Strip the surrounding '{' and '}', and any blanks just inside
        the braces.  */

-    while (*--p == ' ' || *p == '\t')
-      continue;
+    --p;
+  while (c_isspace ((unsigned char) *p))
+      --p;

     p[1] = '\0';
     p[1] = '\0';

-    while (*++decl == ' ' || *decl == '\t')
-      continue;
+    ++decl;
+  while (c_isspace ((unsigned char) *decl))
+      ++decl;

   }
 
   if (! name_start)
   }
 
   if (! name_start)

-    complain_at (loc, _("missing identifier in parameter declaration"));
+    complain (&loc, complaint, _("missing identifier in parameter declaration"));

   else
     {
   else
     {

-      char *name;
-      size_t name_len;
-
-      for (name_len = 1;
-          memchr (alphanum, name_start[name_len], sizeof alphanum);
-          name_len++)
-       continue;
-
-      name = xmalloc (name_len + 1);
-      memcpy (name, name_start, name_len);
-      name[name_len] = '\0';
+      char *name = xmemdup0 (name_start, strspn (name_start, alphanum));

       if (type & param_lex)
         muscle_pair_list_grow ("lex_param", decl, name);
       if (type & param_parse)
       if (type & param_lex)
         muscle_pair_list_grow ("lex_param", decl, name);
       if (type & param_parse)


@@ -780,16 +802,16 @@ version_check (location const *loc, char const *version)
 {
   if (strverscmp (version, PACKAGE_VERSION) > 0)
     {
 {
   if (strverscmp (version, PACKAGE_VERSION) > 0)
     {

-      complain_at (*loc, "require bison %s, but have %s",
-                  version, PACKAGE_VERSION);
-      exit (63);
+      complain (loc, complaint, "require bison %s, but have %s",
+                version, PACKAGE_VERSION);
+      exit (EX_MISMATCH);

     }
 }
 
 static void
 gram_error (location const *loc, char const *msg)
 {
     }
 }
 
 static void
 gram_error (location const *loc, char const *msg)
 {

-  complain_at (*loc, "%s", msg);
+  complain (loc, complaint, "%s", msg);

 }
 
 char const *
 }
 
 char const *